REGULAR ARTICLE
Special Issue onBeyond Classical Chemistry
Excited state dynamics and electron transfer
in a phosphorus(V) porphyrin – TEMPO conjugate
PRASHANTH K PODDUTOORIa,* , NOAH HOLZERa, BRANDON J BAYARDa, YURI E KANDRASHKINb , GARY LIMc, FRANCIS D’SOUZAc
and ART VAN DER ESTd,*
aDepartment of Chemistry & Biochemistry, University of Minnesota Duluth, 1038 University Drive, Duluth, Minnesota 55812, USA
bZavoisky Physical-Technical Institute, FRC Kazan Scientific Center of RAS, Sibirsky Tract 10/7, Kazan 420029, Russian Federation
cDepartment of Chemistry, University of North Texas, 1155 Union Circle, # 305070, Denton, Texas 76203-5017, USA
dDepartment of Chemistry, Brock University, 1812 Sir Isaac Brock Way, St. Catharines, ON L2S 3A1, Canada
E-mail: avde@brocku.ca
MS received 16 March 2021; revised 31 March 2021; accepted 4 May 2021
Abstract. The synthesis and initial spectroscopic characterization of a phosphorus(V) octaethylporphyrin- nitroxide adduct is reported. The nitroxide (2,2,6,6-tetramethylpiperidin-1-yl)oxyl (TEMPO) is axially and covalently bound to the phosphorus porphyrin (PPor?) to give the studied PPor?-TEMPO conjugate. The complex allows the influence of the type of bonding and geometry of the central element in the porphyrin on the excited state dynamics to be investigated. In addition, the high oxidation potential of the PPor?1.35 V vs SCE means that it is a strong oxidant in its excited state and is energetically well-placed to allow charge transfer (CT) from TEMPO to PPor?to occur. Cyclic voltammetry data and DFT computations show that the CT state is the lowest excited state, lying well below the excited sing-doublet and trip-quartet states of the complex. Time-resolved optical data show that the presence of the bound nitroxide leads to a large decrease in the excited state lifetime in acetonitrile, suggesting that rapid charge separation and recombination occurs.
However, transient EPR (TREPR) data reveal that the trip-quartet state is also populated and that emissive spin polarization is generated in the ground state at room temperature, consistent with radical-quartet pair interactions analogous to the radical-triplet pair mechanism. At low temperature, the TREPR data show that the trip-quartet state is formed with strong absorptive net polarization. The shape of the spectrum provides evidence of significant dynamics in the excited state probably due to reversible transitions between the thermally accessible quartet and doublet states.
1. Introduction
Supramolecular complexes consisting of one or more chromophores appended with different electron and energy donors and acceptors or paramagnetic species have attracted a tremendous amount of attention over the past several decades for their potential use in solar energy conversion,1–3 spintronics,4 quantum informa- tion science,5,6and magnetic resonance methods.7,8In particular, there have been many studies of the effect of tethered radicals on the excited state dynamics of
organic chromophores.9–24 In these systems the pres- ence of the additional spin on the radical changes the multiplicity of the states of the chromophore so that the singlet states become sing-doublets (S=1/2) and the triplet states become trip-doublets (S=1/2) and trip- quartets (S=3/2). Decay of the excited sing-doublet to the trip-doublet is spin allowed and promoted by mixing of these two states by the exchange interac- tions between the three unpaired electrons. In most cases, the trip-quartet state is the lowest excited state and its spin sublevels are selectively populated during an intersystem crossing from the sing-doublet and/or trip-doublet resulting in both net and multiplet
*For correspondence
J. Chem. Sci. (2021) 133:65 ÓIndian Academy of Sciences https://doi.org/10.1007/s12039-021-01925-2Sadhana(0123456789().,-volV)FT3](0123456789().,-volV)
polarization.12,15,25 In the majority of these systems, the chromophore has a ring structure with extended p-conjugation and the radical is attached to the chro- mophore via covalent bonds on the periphery of the ring. However, there are a few examples10,12,16,26 in which the radical is attached by either coordination or covalently to a metal or metalloid center via an axial bond perpendicular to the chromophore plane. The spin polarization patterns obtained with axially bound radicals differ from those obtained when the radical is in the same plane as the chromophore and generally display much stronger net polarization.
Recently, we reported the behavior of an alu- minum(III) porphyrin (AlPor) complex in which a TEMPO nitroxide radical was covalently boundviaan axial bond to the aluminum(III) center16 to form AlPor-TEMPO. Under illumination in frozen solution, a purely absorptive spin polarization pattern was observed that was assigned to the excited trip-doublet and trip-quartet states that were proposed to undergo rapid equilibration. Aluminum(III) porphyrins have the unusual property of being able to form both axial covalent and coordination bonds. When the Al(III) center is inserted into the free base porphyrin it assumes a five-coordinate structure with an axial halide or hydroxy group.27 The hydroxy group can then be used to couple porphyrin to substituents by ester formation. However, the five-coordinate Al(III) center is also a strong Lewis acid and can be coordi- nated by Lewis bases such as pyridine and imidazole to form a six-coordinate geometry. In the case of AlPor-TEMPO, coordination of pyridine to the por- phyrin caused a marked change in the spin polarization pattern and a transient electron paramagnetic reso- nance (TREPR) spectrum typical of the lowest excited quartet state with strong multiplet polarization and weak net polarization was observed.16It was proposed that this change in the polarization was caused by changes in the orbital energies and exchange couplings such that the equilibration between the trip-doublet and trip-quartet states becomes slower in the presence of the coordinated pyridine.
Here, we extend our investigation of the role of axial ligands on the excited state dynamics by studying an analogous phosphorus (V) porphyrin complex. In contrast to Al(III) porphyrins, the phosphorus center forms six-coordinate complexes with two axial cova- lent bonds.28 As shown in Scheme 1, insertion of trichloromethylphosophorus into octaethylporphyrin (OEP) results in phosphorus(V)octaethylporphyrin (PPor?) with two covalent axial bonds. The chlorine group can then be displaced to generate PPor?-
TEMPO. The presence of the covalently bound axial methyl group in PPor?-TEMPO compared to coordi- nated pyridine in the case of AlPor-TEMPO allows the influence of the bonding of the axial ligand on the spin polarization to be studied. However, an important additional feature of PPor? is its high oxidation potential compared to other porphyrins.28 This makes the excited state of PPor? a very strong oxidant29–31 and means that excited state decay via electron transfer from TEMPO to PPor? is also possible as has been proposed to occur in fullerene - nitroxide conjugates.32 We will show that the presence of the bound nitroxide leads to a very short excited-state lifetime in polar acetonitrile. However, relaxation of the molecule into its lowest trip-quartet state is observed in a liquid crystalline solvent at room temperature and in an ethanol glass at low temperature. Similar to the AlPor- TEMPO case, the spin polarization patterns show evidence of spin dynamics in the trip-quartet state that leads to net absorptive polarization and partial aver- aging of the zero-field splitting.
2. Experimental 2.1 General
All chemicals and solvents used in this study were purchased from Sigma-Aldrich or Alfa-Aesar. The OEP and chromatography material (Al2O3neutral AW 3) were obtained from Sigma-Aldrich. The precursor porphyrin (PPor-Cl.Cl) and reference porphyrin (PPor- OMe.PF6) were synthesized from OEP by established methods.30Care was taken to avoid the entry of direct, ambient light into the samples in all the spectroscopic and electrochemical experiments. Unless otherwise specified, all the experiments were carried out at 293
±3 K.
Synthesis of PPor-TEMPO.PF6: A solution of OEP (100 mg, 0.187 mmol), CH3PCl2 (50 lL, 0.565 mmol) and 2,6-lutidine (100 lL, 0.859 mmol) in dry dichloromethane (10 mL) was refluxed for 36 h under nitrogen. The solvent was evaporated, and the solute was dried under a vacuum. The obtained crude PPor- Cl.Cl was added a solution of OH-TEMPO (100 mg, 0.580 mmol) in dry dichloromethane (10 mL). The resulting solution was stirred at room temperature for 18 h under a nitrogen atmosphere. The solvent was evaporated, and the residue was chromatographed on neutral alumina. The crude product was loaded onto the column using a benzene: acetonitrile (95:5) solvent mixture and eluted with benzene: methanol (95:5) to
remove the low-polarity impurities. The polarity was then increased by eluting with benzene: methanol (90:10) to collect the desired compound as the chloride salt. To exchange the counterion, the chloride salt was dissolved in 10 mL of ethanol and precipitated with a saturated aqueous solution of NH4PF6. The precipitate was collected by filtration and washed with water and airdried. At this point, the minor impurities (TEMPO, PPor?-OH) found in the product were separated by silica gel column chromatography using dichlor- omethane:ethyl acetate (95:5) as the eluent. Yield:
110 mg (66%). MS (FAB): m/z 749.4788 [M-PF6]? (calcd. 749.4792 for C46H64N5O2P).
2.2 Physical methods
NMR spectroscopy and Mass spectroscopy. NMR spectra were recorded with Bruker Avance 400 MHz spectrometer using CD3CN as the solvent. ESI mass spectra were recorded on a Bruker MicroTOF-III mass spectrometer.
UV-visible absorption and fluorescence spec- troscopy. Steady-state UV-visible absorption spectra of PPor-TEMPO.PF6 and its reference compounds in acetonitrile were recorded with a Cary 100 UV-VIS spectrometer. The concentrations of the samples used for these measurements ranged from 10-6 M (por- phyrin Soret band) to 10-5 M (Q-bands bands) solu- tions. Steady-state fluorescence spectra were recorded using a Photon Technologies International Quanta Master 8075-11 spectrofluorometer, equipped with a
75 W xenon lamp, running FelixGX software. An excitation wavelength of 550 nm, exclusively por- phyrin, was used and the concentrations were held constant for all the compounds.
Electrochemistry. Cyclic and differential pulse voltammetric experiments (CH3CN, 0.1 M tetra- butylammonium hexafluorophosphate, (TBA.PF6)) were performed on a BAS Epsilon electrochemical analyzer (working electrode: glassy carbon, auxiliary electrodes: Pt wire, reference electrode: Ag wire).
The ferrocene couple (E1/2 (Fc?/Fc) = 0.40 V in CH3CN, 0.1M TBA.PF6 under our experimental conditions) was used to calibrate the redox potential values.
Femtosecond transient absorption spectroscopy.
Experiments were performed using an Ultrafast Fem- tosecond Laser Source (Libra) by Coherent incorpo- rating diode-pumped, mode-locked Ti:Sapphire laser (Vitesse) and diode-pumped intracavity doubled Nd:YLF laser (Evolution) to generate a compressed laser output of 1.45 W. For optical detection, a Helios transient absorption spectrometer coupled with a femtosecond harmonics generator both provided by Ultrafast Systems LLC was used. The source for the pump and probe pulses were derived from the funda- mental output of Libra (Compressed output 1.45 W, pulse width 100 fs) at a repetition rate of 1 kHz. 95%
of the fundamental output of the laser was introduced into a harmonic generator which produces second and third harmonics of 400 and 267 nm besides the fun- damental 800 nm for excitation, while the rest of the output was used for the generation of a white light Scheme 1. Synthesis of the investigated PPor?derivatives compounds. Reaction conditions: (i) CH3PCl3, 2,6-lutidine, dichloromethane, reflux under N2 for 36 h, (ii) Dry methanol, stirred for 2 h under N2, (iii) 4-Hydroxy-2,2,6,6- tetramethylpiperidine 1-oxyl, (OH-TEMPO), dry CH2Cl2, stirred for 18 h under N2at room temperature, and (iv) saturated solution of NH4PF6in water.
J. Chem. Sci. (2021) 133:65 Page 3 of 13 65
continuum. In the present study, the second harmonic 400 nm excitation pump was used in all the experi- ments. Kinetic traces at appropriate wavelengths were assembled from the time-resolved spectral data. Data analysis was performed using Surface Xplorer soft- ware supplied by Ultrafast Systems. All measurements were conducted in degassed acetonitrile solutions at 298 K.
Transient EPR Spectroscopy. Transient EPR mea- surements were carried out at room temperature and 80 K using a modified Bruker EPR 200D-SRC X-band spectrometer (Bruker Canada, Milton ON, Canada).
The samples were excited at 532 nm using 10 ns pulses from a Nd:YAG laser at a repetition rate of 10 Hz. For the room temperature experiments the samples were measured in the liquid crystal 4-pentyl- 40-cyanobiphenyl (5CB). The partial ordering of the molecules in the liquid crystal prevents complete averaging of the zero-field splitting allowing the con- tributions from the different spin states to be more easily distinguished from one another. For the low- temperature experiments ethanol was chosen as the solvent to ensure the formation of a glass and to pre- vent any aggregation during the freezing process.
Computations. Density functional theory computa- tions were carried out using Orca 4.2.1.33,34 The geometry was optimized using the unrestricted Kohn- Sham method with B3LYP functional and the 6-31G(d,p) basis set. The resolution of identity and chain of spheres approximations (RIJCOSX)35 were used during the optimization. The influence of the acetonitrile solvent has been taken into account using the conductor-like polarizable continuum model.36
3. Results and discussion
3.1 Synthesis and characterization
The investigated compound was synthesized by exploiting the reactivity of the axial P–Cl bond of PPor?-Cl. The reaction of PPor-Cl.Cl with OH- TEMPO leads to a pairing of the TEMPO and por- phyrin units. After purification and anion exchange, the desired product, PPor-TEMPO.PF6 was obtained in moderate yields. Preliminary characterization of PPor-TEMPO.PF6 was carried out by ESI High Res- olution mass spectrometry. The mass spectrum of the dyad showed peak which correspond to the mass (m/z) [M – PF6]?, for details see the experimental section, and Scheme 1.
DFT calculations. Figure 1 shows the frontier orbitals and charge transfer difference density of the
complex as predicted by DFT computations. The highest doubly occupied orbital (HDOMO) (Fig- ure 1(a)) and lowest unoccupied orbital (LUMO) (Figure 1(b)) of the complex are localized on the PPor? while the singly occupied orbital (SOMO) (Figure 1(c)) is localized on the nitroxide. The com- putations show that there is no significant mixing of the PPor? and nitroxide frontier orbitals and that the two parts of the molecule can be treated as essentially separate entities. The difference density of the lowest excited state calculated by time-dependent DFT is shown in Figure1(d). The red regions localized on the nitroxide indicate the loss of electron density while the green regions localized on the porphyrin are an increase in electron density. Thus, the lowest excited state is a charge transfer state in which the nitroxide has been oxidized by the porphyrin. The calculated ground-state energies of the HDOMO, HDOMO–1, LUMO and LUMO?1 in acetonitrile are given in Table1along with the energies of the excited doublet and quartet states.
The calculations also predict that the HDOMO and HDOMO–1 are nearly degenerate with the HOMO–1 lying only 670 cm-1 below the HDOMO. Similarly, the LUMO?1 is only 104 cm-1 above the LUMO.
Together the HDOMO–1. HDOMO, LUMO and LUMO?1 form the basis for Gouterman’s four orbital model of porphyrins37 that predicts two degenerate allowed transitions referred to as the Soret or B-bands in the blue of the visible spectrum and two degenerate Q-bands in the red. Energetically, the SOMO is also well-separated from the HDOMO and LUMO lying 0.415 eV above the HDOMO and 2.49 eV below the LUMO.
The state energies obtained from time-dependent DFT calculations are also shown in Table 1. As expected, the lowest excited doublet state is predicted to be the CT state, which is nearly doubly degenerate due to the approximate axial symmetry of the por- phyrin and degeneracy of the corresponding orbitals.
The CT state is predicted to lie 1.18 eV above the ground state. The excited sing-doublet state involving excitation of the porphyrin is also doubly degenerate and lies 2.3 eV above the ground state, while the lowest quartet state is at 1.85 eV.
UV2visible absorption spectroscopy. The UV-visible spectra of PPor-TEMPO.PF6 and its ref- erence compounds (PPor-OMe.PF6 and OH-TEMPO) were measured in acetonitrile and the spectra are shown in Figure2(a). The band positions (Q-band and B- or Soret band) and their molar extinction coeffi- cients are summarized in Table 2. As shown in Fig- ure 2(a), the absorption spectrum of the dyad is
essentially a linear combination of its reference com- pounds and the positions and molar extinction coeffi- cients (e) of the porphyrin bands do not change significantly in the dyad compared to the correspond- ing monomer porphyrins. Overall, the absorption studies suggest that any ground state interaction between the porphyrin (PPor?) and the linked TEMPO is sufficiently weak that it causes no discernable change in their electronic structures. The absorption bands of PPor?at 550 and 410 nm have been chosen to excite the PPor? for steady-state fluorescence and transient absorption studies.
Electrochemistry and energetics. Cyclic voltam- metry measurements of the newly synthesized com- pounds were performed in CH3CN with 0.1 M
tetrabutylammonium hexafluorophosphate (TBA.PF6) as the supporting electrolyte and ferrocene as an internal standard. The cyclic voltammogram of PPor- TEMPO.PF6 and its reference compounds PPor- OMe.PF6 and TEMPO derivative (OH-TEMPO) are shown in Figure 3, and the oxidation and reduction midpoint potentials are summarized in Table 2. The redox processes of all of the compounds are found to be one-electron reversible processes based on the peak-to-peak separation values, and the cathodic-to- anodic peak current ratio. The observed voltammo- gram of PPor-TEMPO.PF6 is assigned to the com- bination of PPor-OMe.PF6 and OH-TEMPO derivative. During the cathodic scan, the dyad PPor- TEMPO.PF6 showed two reduction processes, which we assign to the first and second reduction of the PPor? moiety. The anodic scan of the dyad reveals two oxidation processes, which are assigned to the first oxidation of TEMPO and first oxidation of PPor?unit. The oxidation potential of the TEMPO is shifted slightly more positive due to the electron withdrawing ability of the PPor? unit. Comparison of the dyads and their reference compounds shows that the redox processes are the sum of those of the reference compounds PPor-OMe.PF6 and OH- TEMPO, which again indicates that the porphyrin Figure 1. Frontier molecular orbitals and charge transfer state difference density in PPor?-TEMPO. (a) Highest doubly occupied orbital (b) Lowest unoccupied orbital (c) singly occupied orbital (d) CT state density difference. The green surfaces are positive and red surfaces are negative. The orbitals and difference density were calculated using the program Orca33,34and the surfaces were rendered using Avogadro.38
Table 1. Computed orbital and doublet state energies of PPor?-TEMPO in eV.
Orbital Orbital energy State State energy LUMO?1 -2.776 Sing-doublet 2.355, 2.365 LUMO -2.789 Trip-quartet 1.851, 2.038
SOMO -5.282 CT State 1.179, 1.189
HDOMO -5.697 Ground state 0
HDOMO-1 -5.780
J. Chem. Sci. (2021) 133:65 Page 5 of 13 65
Figure 2. (a) Absorption spectra of PPor-OMe.PF6 (green), OH-TEMPO (red), and PPor-TEMPO.PF6 (purple) in acetonitrile. (b) Fluorescence spectra of PPor-OMe.PF6(green), physical mixture of OH-TEMPO?PPor-OMe.PF6(red), and PPor-TEMPO.PF6 (purple) in acetonitrile. Excitation wavelength was used 550 nm. All the sample concentrations maintained at 10lM.
Table 2. Redox and optical data of investigated compounds in acetonitrile.
Sample
Redox midpoint potentials (V)
Absorption data kmax, nm (loge)
Fluorescence data kmax, nm (%Q) Oxidation Reduction
PPor-OMe.PF6 1.35 -0.81,-1.26 590 (4.10), 549 (4.07), 416 (5.38), 355 (4.38) 596, 652 PPor-TEMPO.PF6 0.77, 1.44 -0.82,-1.32 586 (4.06), 544 (4.09), 416 (5.40), 357 (4.43) 592, 648 (87)
OH-TEMPO 0.70 – 241 (3.26) -
Figure 3. Left: Cyclic voltamogramas of OH-TEMPO (red), PPor-OMe.PF6(green), and PPor-TEMPO.PF6(purple) in acetonitrile with 0.1 M TBA.PF6. Right: Energy level diagram of PPor-TEMPO.PF6in acetonitrile. The possible electron transfer (ET), internal conversion (IC) and intersystem crossing (ISC) processes are shown. The energies of the sing- doublet, trip-quartet and trip-doublet states (2.09 eV and 1.65 eV) are the literature values29of the PPor?excited singlet and triplet states and do not include exchange interaction between the nitroxide and the porphyrin.
electronic structure is not perturbed by presence of TEMPO.
The redox potentials can be used in combination with optical data to estimate the energetics of the electron transfer reactions in the PPor-TEMPO.PF6. The energy of the charge-separated state ECS (rela- tive to the ground state) and the free-energy changes for charge-separation (DGCS) can be calculated using the Rehm and Weller equations. Because the mag- nitude of the exchange interaction between the nitroxide and the porphyrin is difficult to determine experimentally, the energies of the sing-doublet, trip- doublet and trip-quartet states have been taken as the literature values of first excited singlet-state and triplet-state energies (E0-0 = 2.09 eV and Etriplet
=1.65 eV)29 of PPor?. The ECS is the energy of the charge-separated state and is the difference between the first oxidation potential of the donor TEMPO and first reduction potential of the acceptor PPor? units.
DGCS ¼ECSE00 ð1Þ
ECS ¼E1ox=2ðTEMPOÞ Ered1=2ðPPorþÞ þGs ð2Þ
GS¼ e2 4pe0RDAeS
ð3Þ
where RD-A is the center-to-center distance between the donor (TEMPO) and the acceptor (PPor?), and has been taken as the distance between the phosphorus atom and the center of the nitroxide N–O bond (6.35 A˚ ) using the optimized geometry from the DFT calculations.eSis the dielectric constant of the solvent used for the photophysical and redox potential mea- surements, in this case acetonitrile. The Gs value in acetonitrile was found to be-0.0605 eV. Using these equations, the energy of the charge-separated state PPor-TEMPO? is found to be 1.53 eV, which lies well below the excited sing-doublet, trip-doublet and trip-quartet state energies showing that PPor? is energetically well suited to act as a photosensitizer for electron transfer from TEMPO to the 1(PPor?)* unit.
With the above information, the energy level diagram was constructed as shown in Figure 3.
Figure 4. Transient absorption spectra of PPor-OMe.PF6 in acetonitrile. (a) Femtosecond absorbance changes at the indicated times. (b) Time dependent absorbance change at 1185 nm. (c) Nanosecond absorbance changes at the indicated times (d) Time dependent absorbance change at 465 nm.
J. Chem. Sci. (2021) 133:65 Page 7 of 13 65
Qualitatively, the state energies predicted by the DFT calculations (Table 1) agree fairly well with the experimental values, although the energy difference between the trip-quartet state and CT state is signifi- cantly overestimated in the calculations.
Fluorescence Spectroscopy. The steady-state fluo- rescence spectra of PPor-TEMPO.PF6 and its refer- ence compound PPor-OMe.PF6 in acetonitrile are shown in Figure2(b). The spectra have been measured with excitation at 550 nm, where the absorption is exclusively due to PPor?. The spectral shapes and the emission maxima of the dyad are essentially the same as its reference compound, see Table 1. However, the fluorescence of the dyads is strongly quenched (87%) compared to that of the reference compound. The observed fluorescence quenching could be assigned to two possible intramolecular quenching mechanisms.
One of them would be reductive electron transfer from TEMPO to the singlet excited state of the PPor? (1(PPor?)*) as the anticipated charge-separated state is well placed with sufficient driving force (DGCS = -0.56 eV). The second possibility would be fast internal conversion (IC) to the trip-doublet state in the presence of the free-radical TEMPO. Additionally, intermolecular quenching was also tested by measur- ing the fluorescence spectra of PPor-OMe.PF6and the 1: 1 physical mixture PPor-OMe.PF6 and OH- TEMPO, see Figure 2(b). As shown the fluorescence intensity is identical to the reference compound, which rules out the intermolecular mechanism. The exchange coupling between the unpaired electron on TEMPO and the electrons in the HDOMO and LUMO of the porphyrin splits the 3(PPor?)* state into trip-doublet
and trip-quartet states as shown in Figure 3. The rel- ative energies of the trip-doublet and trip-quartet is determined by the exchange coupling between the triplet and doublet spins and the trip-quartet is expected to be lower in energy than the trip-doublet. If the coupling between the nitroxide spin and the two unpaired electrons on the porphyrin is different the sing-doublet state becomes mixed with the trip-dou- blet state and fast relaxation occurs.39The trip-doublet can then relax by spin-orbit coupling mediated inter- system crossing (ISC) to the trip quartet or the ground state. The trip-doublet state may also decay further to the charge separated state, but this process is spin forbidden from the trip-quartet state. Thus, the relative populations of the trip-quartet and CT states depend on the relative rates of the relaxation processes shown in Figure3.
4. Time-resolved optical spectroscopy
Figure 4 shows pump-probe measurements of PPor?- OMe in two different time regions. Figures 4(a) and 4(b) show the absorbance changes in the picosecond to nanosecond time range, Figures 4(c) and 4(d) show corresponding data in the microsecond range. In both time regimes the negative peaks at 549 and 590 nm are due to the bleaching of the Q-bands when the molecule is excited. The weaker negative peak at *640 nm arises from the porphyrin fluorescence. The strong absorbance increase band at *440 nm and the weak broad band between 1100 and 1300 nm are due to excited state absorbance. The upper time trace is the absorbance change at 1185 nm and represents the excited singlet state absorbance that decays with a lifetime of 5.25 ns31as expected for the excited singlet state of a porphyrin.
In Figure4(c) the absorption difference spectrum in the microsecond time region is shown. In this time range, the bleaching of the Q-bands is still observed but the negative fluorescence band at*640 nm seen at an early time is no longer present. Instead, a broad absorption increase band between *640 nm and 800 nm from the excited triplet state absorbance is seen.
The time trace in the bottom right of Figure 4 shows the decay of the absorbance difference band at 465 nm. The red curve is a fit of the data with three exponentials with lifetimes of 8, 20 and 37ls. Thus, as expected the data shows that in PPor?-OMe excitation leads to decay of the excited singlet state via fluores- cence and intersystem crossing (ISC). The ISC life- time is on the order of a few ns and the triplet state has a lifetime of a few tens of microseconds.
Figure 5. Femtosecond transient absorbance spectra of PPor-TEMPO.PF6in acetonitrile.
Figure 5 shows corresponding absorbance changes for PPor?-TEMPO in the picosecond time regime.
These spectra reveal the significant impact that the presence of nitroxide has on the excited state dynam- ics. First, all of the absorption changes including the Q-band bleaching peaks decay nearly to zero within the 3 ns time window of the experiment. Thus, the overall excited state lifetime is much shorter. Second, the emissive band at *640 nm due to the porphyrin fluorescence is barely visible underneath the excited state absorbance, again showing that the excited sin- glet state lifetime is much shorter. The decrease in the excited state lifetime could be due to the presence of the unpaired electron spin on the nitroxide and/or electron transfer from the nitroxide to the excited porphyrin. To better understand what lead to the absorbance changes that can be expected from the electron transfer, spectroelectrochemistry measure- ments of the reduction of PPor?-OMe are useful.
Although the charge-transfer state of PPor?-TEMPO differs from the reduced form of the porphyrin, the spectra of the two species are expected to be similar.
The UV/Vis spectra of PPor?-OMe as it undergoes its first reduction has been reported previously31 and show that the reduction leads to the loss of the Soret band at 416 nm and the two Q-bands at 549 and 590 nm, while new bands at 447 nm and 750 nm appear.
Thus, we can expect to observe an increase in the absorbance at these two wavelengths in pump-probe measurements of PPor?-TEMPO if the CT is being Figure 6. Room temperature TREPR spectra in the liquid crystal 5CB at different delay times. (a): PPor?-OMe (blue) and an equimolar mixture of PPor?-OMe and OH-TEMPO (red). The top and bottom spectra are at 200 ns and 750 ns after the laser flash. (b): Corresponding spectra of PPor?-TEMPO at 70 ns and 750 ns after the laser flash.
Figure 7. Low temperature TREPR spectra PPor?- TEMPO (black spectrum) and PPor?-OMe (red spectrum) in ethanol glass at 80 K. The spectra were extracted from the time/field datasets in a time window centred at 800 ns after the laser flash.
J. Chem. Sci. (2021) 133:65 Page 9 of 13 65
formed. In the spectra of the reference compound in Figure 3, there are also absorbance increase bands at both of these wavelengths, thus it is difficult to dis- tinguish between the porphyrin excited states and the reduced porphyrin. Nonetheless, the rapid decay of the PPor?-TEMPO absorbance changes and the observed absorbance increases around 450 nm and 750 nm suggest that the charge transfer state may be involved.
5. Transient electron paramagnetic resonance spectroscopy (TREPR)
Figure6shows a comparison of the room temperature transient EPR spectra of PPor?-OMe (blue spectra), an equimolar mixture of PPor?-OMe and OH-TEMPO (red spectra) and PPor?-TEMPO (black spectra) in the liquid crystal 5CB. In Figure6(a), the top spectra were extracted from the time/field dataset in a time window centred at 200 ns after the laser flash, while the bottom spectra are at 750 ns. The early spectra of PPor?-OMe and the physical mixture of PPor?-OMe and OH- TEMPO are identical and show an emission (E)/ab- sorption (A) polarization pattern and arises from the triplet state of the porphyrin. In the nematic phase of the liquid crystal, the molecules undergo fast aniso- tropic tumbling and hence, the zero-field splitting of the triplet state is not averaged to zero but a non-zero value that is determined by the order parameters of the porphyrin. As is apparent in Figure 6(a) (top), the emissive peak at *337 mT is more intense than the absorptive peak at*357 mT. This net polarization of the triplet state probably arises from its dynamics. The porphyrin has two nearly degenerate excited triplet states associated with the Qxand Qyabsorption bands.
The calculated orbital and state energies of PPor?(not shown), suggest that these two states are separated by an energy difference of a few hundred wavenumbers, hence some equilibration between them can be expected at room temperature. We have shown recently that such transitions lead to net polarization of the triplet state.40 However, the effect of the equili- bration of the two triplet states is also similar to that caused by anisotropic motion, so that it is possible that the molecular motion may contribute to the net polarization.
At 750 ns after the laser flash, the spectrum of PPor?-OMe (Figure 6(a), blue spectrum, bottom) retains the same polarization pattern but the intensity of the spectrum decays as a result of equalization of the populations of the spin sublevels due spin-lattice (T1) relaxation. In contrast, in the presence of OH- TEMPO, strong emissive features from the nitroxide,
appear in the middle of the spectrum and the polar- ization of the triplet state changes. The emissive polarization of the nitroxide occurs as a result of col- lisions between the nitroxide and triplet state and can be explained with the radical-triplet pair mechanism (RTPM).26,41–45The change in the polarization of the high-field part of the triplet spectrum is due to the depopulation of the triplet sublevels that interact most strongly with the radical. The hyperfine component of the nitroxide at the highest field is also much more intense than the other two peaks. This is partly due to the narrower linewidth of the peak as a result of incomplete anisotropic motional averaging. However, it also appears to have stronger spin polarization. This implies that the doublet-triplet mixing that leads to polarization of the radical may be dependent on the orientation of the nitroxide nitrogen nuclear spin.
Figure6(b) shows corresponding spectra for PPor?- TEMPO at 70 ns (upper black spectrum) and 750 ns (lower black spectrum) after the laser flash. The early spectrum shows broad absorptive features about 20 mT wide centred around a narrow absorptive peak.
The width of the broad features is too large to be associated with any of the possible doublet states of PPor?-TEMPO and we assign them to the ±3/2 $
±1/2 transitions of the lowest trip-quartet state. The narrow peak in the centre of the spectrum is assigned to the ?1/2 $ -1/2 transition of the quartet state.
However, it could also potentially contain contribu- tions from the trip-doublet and ground doublet and charge-transfer doublet states. At a late time, an emissive signal from the ground-state is observed. We postulate that the emissive polarization is generated by collisions between molecules in the excited quartet state and ground state in analogy to the polarization generated by collisions between the triplet and doublet states in the physical mixture. The linewidths of the peaks in the ground state spectrum are broader than in the physical mixture as a result of slower tumbling of the molecule due to its larger size. The relative heights of the three hyperfine components are also different than in the physical mixture. This is at least partly due to the fact that the ordering and rotational correlation time of PPor?-TEMPO in the liquid crystal environ- ment is different than that of OH-TEMPO. It is also possible that the polarization of the nitroxide shows different selectivity for the hyperfine components than in the physical mixture. However, the most important observation is that the trip-quartet state of PPor?- TEMPO decays with a lifetime on the order of a few hundred ns and only the ground state is observed at a longer time.
Figure 7 shows a comparison of the low-tempera- ture TREPR spectra of PPor?-TEMPO and PPor?- OMe in ethanol glass at 80 K. The spectrum of PPor?- OMe is typical of the triplet state of a porphyrin populated by spin-orbit coupling mediated intersystem crossing and is essentially identical to the spectrum reported recently in 2-methyltetrahydrofuran.46 The spectrum of PPor?-TEMPO is purely absorptive and is significantly narrower than that of the PPor?-OMe.
We assign the spectrum to the trip-quartet state of the complex based on its width and general shape. Inter- system crossing to the trip-quartet state of coupled quartet-doublet systems generally leads to a polariza- tion pattern in which the wings of the spectrum show predominant multiplet polarization (E/A or A/E) with some net polarization that results in an absorptive or emissive central peak.10–12,15,18 However, in cases in which there are nearby quartet and doublet states that are thermally accessible and/or the exchange interac- tion between the triplet and doublet states is compa- rable to the Zeeman energy net polarization can develop.16,25,40,47,48
Here, we postulate that equilibra- tion between the low-lying trip-doublet and trip- quartet states leads to partial averaging of the zero- field splitting tensor and generates the observed pre- dominant absorptive polarization.
6. Conclusions
The fast decay observed in the femtosecond transient absorbance data suggests that in acetonitrile the sing- doublet state may decay by electron transfer from the nitroxide to PPor?. This process is not spin selective.
Thus, the CT state should not contribute to the TREPR signals if it is populated via this pathway. However, the optical data do not provide unambiguous evidence for ET from the sing-doublet and further studies are in progress to investigate whether it occurs or not. The broad wings of the room temperature TREPR spectra show clear evidence for the formation of the quartet state. Taken together with the observed fluorescence quenching it suggests that IC to the trip-doublet state and ISC to the trip-quartet state compete favourably with ET. This is somewhat surprising since the mixing of the sing-doublet and trip-doublet states induced by the exchange coupling between the triplet and doublet spins should be relatively weak and therefore is not expected to have a strong influence on the depopula- tion of the excited sing-doublet state.
In contrast to ET from the sing-doublet, decay of the trip-doublet to the CT state could generate spin polarization and would retain any polarization of the
trip-doublet state.32 Thus, it may contribute to the absorptive central peak in the TREPR spectrum of PPor?-TEMPO observed at an early time (Fig- ure6(b)). In addition, any polarization of the CT state should be transferred to the ground state during reverse ET. At room temperature, it is possible that some of the ground state polarization observed at late times is generated in this way. At low temperature (Figure 7) the spectrum does not show any features that can be assigned to the ground state. Since the trip- quartet state displays strong net polarization and the trip-doublet state can probably be thermally populated from the trip-quartet, decay to the ground state either directly from the trip-quartet or via the trip-doublet and/or CT state appears to be slow.
The very strong absorptive net polarization and the broadening of the wings of the low-temperature trip- quartet state spectrum suggest that the excited state undergoes rapid dynamics. The near degeneracy of the two lowest porphyrin triplet states means that transitions between them can occur and depending on how large the exchange coupling is between the tri- plet and doublet spins the trip-doublet may also be thermally accessible from the trip-quartet. This cre- ates a rather complex situation in which four (par- tially) averaged spectra can contribute to the observed signal.
To better understand the excited state dynamics in this system we are extending these studies to investigate the dependence of the polarization on the length of the bridge between porphyrin and the nitroxide.
Acknowledgements
This work was supported by the Grant-in-Aid program from the University of Minnesota Duluth to PPK, the Natural Sciences and Engineering Research Council Canada (Dis- covery Grant 2015-04021 to AvdE) and the National Science Foundation (Grant No. 2000988 to FD).
References
1. Maddah HA, Berry V and Behura SK 2020 Biomolec- ular photosensitizers for dye-sensitized solar cells:
Recent developments and critical insights Renew.
Sustain. Energy Rev. 121109678
2. Fukuzumi S, Lee Y-M and Nam W 2018 Mimicry and functions of photosynthetic reaction centers Biochem.
Soc. Trans.461279
3. Mora SJ, Odella E, Moore GF, Gust D, Moore TA and Moore AL 2018 Proton-Coupled Electron Transfer in Artificial Photosynthetic Systems Acc. Chem. Res. 51 445
J. Chem. Sci. (2021) 133:65 Page 11 of 13 65
4. Sanvito S 2011 Molecular spintronicsChem. Soc. Rev.
403336
5. Bruss D and Leuchs G 2019 Quantum Information:
From Foundations to Quantum Technology Applica- tions(Weinheim, Germany: John Wiley)
6. Rugg BK, Krzyaniak MD, Phelan BT, Ratner MA, Young RM and Wasielewski MR 2019 Photodriven quantum teleportation of an electron spin state in a covalent donor–acceptor–radical systemNat. Chem.11 981
7. Tripathi A and Rane V 2019 Toward Achieving the Theoretical Limit of Electron Spin Polarization in Covalently Linked Radical-Chromophore Dyads J.
Phys. Chem. B123 6830
8. Liu G, Liou S-H, Enkin N, Tkach I and Bennati M 2017 Photo-induced radical polarization and liquid-state dynamic nuclear polarization using fullerene nitroxide derivativesPhys. Chem. Chem. Phys.1931823 9. Fujisawa J, Ishii K, Ohba Y, Iwaizumi M and Yamauchi
S 1995 Electron spin polarization transfer from excited triplet porphyrins to a nitroxide radical via a spin exchange mechanismJ. Phys. Chem.9917082 10. Ishii K, Hirose Y, Fujitsuka H, Ito O and Kobayashi N
2001 Time-resolved EPR, fluorescence, and transient absorption studies on phthalocyaninatosilicon cova- lently linked to one or two TEMPO radicals J. Am.
Chem. Soc.123702
11. Ishii K, Ishizaki T and Kobayashi N 1999 Experimental evidence for a selection rule of intersystem crossing to the excited quartet states: Metallophthalocyanines coor- dinated by 4-amino-TEMPOJ. Phys. Chem. A1036060 12. Tarasov VF, Islam SS, Ohba Y, Forbes MD and Yamauchi S 2011 Multifrequency TREPR Investigation of Excited-State ZnTPP/Nitroxide Radical Complexes Appl. Magn. Reson.41175
13. Chernick ET, Casillas R, Zirzlmeier J, Gardner DM, Gruber M, Kropp H, et al. 2015 Pentacene appended to a TEMPO stable free radical: the effect of magnetic exchange coupling on photoexcited pentacene J. Am.
Chem. Soc.137857
14. Teki Y 2020 Excited-State Dynamics of Non-Lumines- cent and Luminescent pi-Radicals Chem.- Eur. J. 26 980
15. Kandrashkin Y and van der Est A 2003 Light-induced electron spin polarization in rigidly linked, strongly coupled triplet–doublet spin pairsChem. Phys. Lett.379 574
16. Poddutoori PK, Kandrashkin YE, Karr P and van der Est A 2019 Electron spin polarization in an Al (III) porphyrin complex with an axially bound nitroxide radicalJ. Chem. Phys.151204303
17. Moons H, Goovaerts E, Gubskaya VP, Nuretdinov IA, Corvaja C and Franco L 2011 W-band transient EPR and photoinduced absorption on spin-labeled fullerene derivativesPhys. Chem. Chem. Phys.133942
18. Poddutoori PK, Pilkington M, Alberola A, Polo V, Warren JE and van der Est A 2010 Spin- Spin Interactions in Porphyrin-Based Monoverdazyl Radical Hybrid Spin SystemsInorg. Chem.493516
19. Teki Y, Miyamoto S, Iimura K, Nakatsuji M and Miura Y 2000 Intramolecular spin alignment utilizing the excited molecular field between the triplet (S= 1)
excited state and the dangling stable radicals (S= 1/2) as studied by time-resolved electron spin resonance:
observation of the excited quartet (S= 3/2) and quintet (S= 2) states on the purely organic p-conjugated spin systems J. Am. Chem. Soc.122984
20. Teki Y, Kimura M, Narimatsu S, Ohara K and Mukai K 2004 Excited High-Spin Quartet (S= 3/2) State of a Novel p-Conjugated Organic Spin System, Pyrene- Verdazyl RadicalBull. Chem. Soc. Jpn. 7795
21. Takemoto Y and Teki Y 2011 Unique Dynamic Electron-Spin Polarization and Spin Dynamics in the Photoexcited Quartet High-Spin State of an Acceptor- Donor-Radical TriadChemPhysChem12104
22. Nolden O, Fleck N, Lorenzo ER, Wasielewski MR, Schiemann O, Gilch P and Richert S 2020 Excitation energy transfer and exchange-mediated quartet state formation in porphyrin-trityl systems Chem. Eur. J.27 2683
23. Zhang X, Sukhanov AA, Yildiz EA, Kandrashkin YE, Zhao J, Yaglioglu HG and Voronkova VK 2021 Radical-Enhanced Intersystem Crossing in a Bay-Sub- stituted Perylene Bisimide- TEMPO Dyad and the Electron Spin Polarization Dynamics upon Photoexci- tation ChemPhysChem2255
24. Giacobbe EM, Mi Q, Colvin MT, Cohen B, Ramanan C, Scott AM, et al. 2009 Ultrafast intersystem crossing and spin dynamics of photoexcited perylene-3, 4: 9, 10-bis (dicarboximide) covalently linked to a nitroxide radical at fixed distancesJ. Am. Chem. Soc.131 3700
25. Kandrashkin YE and van der Est A 2019 The triplet mechanism of electron spin polarization in moderately coupled triplet-doublet rigid complexes as a source of the enhanced ?1/2$-1/2 transitionsJ. Chem. Phys.
151 184301
26. Fujisawa J, Ohba Y and Yamauchi S 1997 Electron-spin polarizations generated from interactions between excited triplet porphyrins and stable radicals studied by time-resolved electron paramagnetic resonance J.
Phys. Chem. A101434
27. Furhop JH, Kadish KM and Davis DG 1973 Redox behavior of metallo oxtaethylporhyrins J. Am. Chem.
Soc. 955140
28. Akiba K, Nadano R, Satoh W, Yamamoto Y, Nagase S, Ou Z, et al. 2001 Synthesis, structure, electrochemistry, and spectroelectrochemistry of hypervalent Phosphorus (V) octaethylporphyrins and theoretical analysis of the nature of the PO bond in P (OEP)(CH2CH3)(O)Inorg.
Chem.405553
29. Poddutoori PK, Thomsen JM, Milot RL, Sheehan SW, Negre CF, Garapati VK, et al. 2015 Interfacial electron transfer in photoanodes based on phosphorus (v) por- phyrin sensitizers co-deposited on SnO 2 with the Ir (III) Cp* water oxidation precatalystJ. Mater. Chem. A 3 3868
30. Poddutoori PK, Dion A, Yang S, Pilkington M, Wallis JD and van der Est A 2010 Light-induced hole transfer in a hypervalent phosphorus (V) octaethyl- porphyrin bearing an axially linked bis (ethylene- dithio) tetrathiafulvalene J. Porphyr. Phthalocy. 14 178
31. Poddutoori PK, Lim GN, Pilkington M, D’Souza F and van der Est A 2016 Phosphorus (V) Porphyrin-
Manganese (II) Terpyridine Conjugates: Synthesis, Spectroscopy, and Photo-Oxidation Studies on a SnO2 SurfaceInorg. Chem. 5511383
32. Rozenshtein V, Berg A, Stavitski E, Levanon H, Franco L and Corvaja C 2005 Electron spin polarization of functionalized fullerenes. Reversed quartet mechanism J. Phys. Chem. A10911144
33. Neese F 2012 The ORCA program system Wiley Interdiscip. Rev. Comput. Mol. Sci.273
34. Neese F 2018 Software update: the ORCA program system, version 4.0 Wiley Interdiscip. Rev. Comput.
Mol. Sci.8e1327
35. Izsa´k R and Neese F 2013 Speeding up spin-compo- nent-scaled third-order pertubation theory with the chain of spheres approximation: the COSX-SCS-MP3 methodMol. Phys.111 1190
36. Barone V and Cossi M 1998 Quantum calculation of molecular energies and energy gradients in solution by a conductor solvent modelJ. Phys. Chem. A1021995 37. Gouterman M, Wagnie`re GH and Snyder LC 1963
Spectra of porphyrins: Part II. Four orbital model J.
Mol. Spectrosc.11108
38. Hanwell MD, Curtis DE, Lonie DC, Vandermeersch T, Zurek E and Hutchison GR 2012 Avogadro: an advanced semantic chemical editor, visualization, and analysis platformJ. Cheminformat.4 17
39. Ake RL and Gouterman M 1969 Porphyrins XIV.
Theory for the luminescent state in VO Co, Cu complexesTheor. Chim. Acta1520
40. Kandrashkin YE, Di Valentin M and van der Est A 2020 Reversible triplet energy hopping in photo-excited molecules: A two-site model for the spin polarization J. Chem. Phys.153094304
41. Blattler C, Jent F and Paul H 1990 A Novel Radical- Triplet Pair Mechanism for Chemically-Induced Elec- tron Polarization (CIDEP) of Free-Radicals in Solution Chem. Phys. Lett.166375
42. Bla¨ttler C and Paul H 1991 CIDEP after laser flash irradiation of benzil in 2-propanol. Electron spin polarization by the radical-triplet pair mechanism Res.
Chem. Intermed. 16201
43. Kobori Y, Takeda K, Tsuji K, Kawai A and Obi K 1998 Exchange interaction in radical-triplet pairs: Evi- dences for CIDEP generation by level crossings in triplet-doublet interactions J. Phys. Chem. A 102 5160
44. Kawai A, Watanabe Y and Shibuya K 2002 Time resolved ESR study on energy difference of quartet and doublet states in radical-triplet encounter pairs Mol.
Phys.100 1225
45. Shushin AI 1993 The relaxational mechanism of net CIDEP generation in triplet—radical quenching Chem.
Phys. Lett.208 173
46. Zarrabi N, Bayard BJ, Seetharaman S, Holzer N, Karr P, Ciuti S, et al. 2021 A charge transfer state induced by strong exciton coupling in a cofacial l-oxo- bridged porphyrin heterodimer Phys. Chem. Chem.
Phys. 23960
47. Kandrashkin YE, Asano MS and van der Est A 2006 Light-induced electron spin polarization in vanadyl octaethylporphyrin: I. Characterization of the excited quartet state J. Phys. Chem. A1109607
48. Kandrashkin YE, Asano MS and van der Est A 2006 Light-induced electron spin polarization in vanadyl octaethylporphyrin: II. Dynamics of the excited statesJ.
Phys. Chem. A1109617
J. Chem. Sci. (2021) 133:65 Page 13 of 13 65
